boltz2-nim
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFE
Full Analysis
- [CREDENTIALS_UNSAFE]: The skill facilitates the use of an API key (NGC_API_KEY) through secure practices, specifically recommending the use of environment variables and local .env files for credential storage and retrieval.
- [EXTERNAL_DOWNLOADS]: The skill provides instructions for downloading a Docker image from nvcr.io. This is the official container registry for the skill's authoring organization (NVIDIA) and is a legitimate, trusted source for the tool's operation.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external biological sequence data and chemical SMILES strings. These ingestion points are limited to scientific analysis by the underlying model and do not expose any dangerous system capabilities to potential injection attacks.
Audit Metadata